Drivers for 811b's ?

[rustnrot]

Tom Lang over in Augusta, GA with a pair of Heathkit AS-101s that I have been listening to for years with an extra piezo tweeter for each one.

Intrigued with this thread (and having read, or tried to read, the many others here and in other forums), let me take a moment to summarize what I have learned and what can be done to improve these speakers.

1. Many different drivers from many different companies will fit the horns in the AS-101s. These horns are Altec 811b.

2. Many agree, as from factory, the 811b horns in the AS-101's lack highs and are bloated in the mids.

3. There has been alot of research here recommends the BMS 4550ND drivers for the 811b horns and there is group buy going on to purchase these in bulk at a discount.

4. To optimize the BMS 4550ND's with the 811b's, a new crossover (with mid frequency attenuation) is required.

Here are my questions, (also please verify the above that I have stated is also correct).

1. Is there a crossover that is an improvement over the factory one in the AS-101 that essentially attenuates the mid freqs and therefore accenuates the highs? Would this be the crossover for the Altec 19?

2. Exactly what and where is the recommended crossover to utilize as part of this group buy exercise for the BMS 4550ND's in the 811b AS-101 horns?

 

[Zilch]

Hello, Tom.

Heath AS-101 is a kit version of Altec Valencia. The 806 compression drivers in them have excessively forward midrange and rolled-off highs. The stock passive crossover has some compensation built into the highpass filter which tames the mids somewhat, however there is insufficent headroom between the woofer and compression driver sensitivities to do adequate full-spectrum compensation.

Flattened response and HF extension out to ~14 kHz may be achieved via biamping the stock drivers and engaging the CD horn compensation built into an active crossover such as the Behringer CX3400. Further extending the VHF response to 18 - 20 kHz is apparently not practical with the 806 driver:

http://audiokarma.org/forums/showthread.php?p=1473345&#post1473345

A better result may be accomplished passively if you're willing to swap out the compression drivers for improved Altecs with higher sensitivity and tangerine phase plugs, and upgrade the crossover, for example, to that used in Model 19 or similar. It may also be possible to supplement the stock crossover with an additional compensation network. These approaches are untested and undocumented presently, though similar projects are in progress.

Alternatively, either BMS 4550 or 4552nd may be substituted for the HF drivers, retaining the stock crossover and adding on a supplemental compensation filter, which upgrade has been tested and documented for both the 8- and 16-Ohm versions of Valencia and similar. One AK member, Steve O, above, is completing an improved, comprehensive crossover redesign for incorporating BMS drivers in these systems.

There are additional options by others, largely undocumented to date, and some indications that Valencia bass response may also be improved.

If you tell me the impedances of your drivers, and/or the capacitor values in your crossovers, I can link you to the requisite supplemental compensation network for the current BMS upgrade.... :yes:

 

[dgwojo]

The group buy is at 10 as of today, 2/15/2008, and they're for 8 ohm 4552ND BMS drivers, not the 4550's, we're going with the more linear neo magnet versions, the drivers are supposed to be back in stock next week. Dave. :yes:

 

[geaugafletcher]

The CX3400 has a provision for time-aligning the LF driver. Is that necessary in this case, or do the drivers' relative positions and reversing the HF polarity take care of it?

In short...I'm looking into the Behringer as well as an old Altec active - the Altec has no time alignment adjustments. It does have a nifty provision for CD compensation, though...

 

[Zilch]

The CX3400 has a provision for time-aligning the LF driver. Is that necessary in this case, or do the drivers' relative positions and reversing the HF polarity take care of it?

I don't know where the acoustic center of a driver on an 811B horn actually is, but Altec seems to have been somewhat schizo about it, claiming in VOTT where the woofer is at the rear, that the centers were aligned with the horn driver back there as well, yet on the other hand claiming that it was correct in Valencia et al. with the woofer at the front.

The only way I can reconcile this is to believe they were talking about phase alignment instead, in the latter case, at the crossover frequency. Wavelength at 800 Hz is 16.95", by my calculation, so, assuming the acoustic center is at the driver mounting flange (it may be further back, at the front plate, or ahead, at the throat terminus, perhaps,) then it might be conceivable that the driver and woofer would be phase aligned in Valencia, but it is clearly not time aligned. See the Altec Tech Note on this, below. Moving the horn back just the depth of its lips is sufficient change in displacement as to require phase reversal of the compression driver.

In any case, time alignment and phase alignment are distinct in concept. The only way for time alignment to occur is if the times of arrival from the woofer and the compression driver are coincident, and that is largely independent of frequency, in the simple view. It's possible to get phase alignment correct for misaligned drivers at a given frequency by shifting the phase (sometimes accomplished simply by reversing the leads as shown in the attachment,) but the only way to get the time alignment correct is via delay of the sound from the source closest the listener. Generally, if time alignment is correct, phase alignment is, as well....

 

[rustnrot]

Zilch, my existing crossover is factory Figure 3 attached .pdf. Drivers are 8 ohm. Please link me to correct crossover mods for group buy 4552ND drivers.

A few more quick questions:

1. what are " tangerine phase plugs"

2. why did dgwojo reference the price of 4540ND's in post #61? The group buy is for 4552ND's, correct?

3. there is mention of other drivers such as those manufactured by Selenium and others are mentioned in this thread. Is it a fair statement that the most-tested-documented mod is the 4552ND replacment and the others, while they may be viable alternatives are not thoroughly tested yet?

 

[dgwojo]

2. why did dgwojo reference the price of 4540ND's in post #61? The group buy is for 4552ND's, correct?

The pair of threaded 4540's are for Zilch only, yes the 4552's are what we're going for in the group, all drivers will be 8 ohms, we're only in need of two more to make the dozen quantity needed for the discount. Dave.

 

[dgwojo]

A few more quick questions:

1. what are " tangerine phase plugs"

From 30 years ago and still produced today, they replace the old metal phase plugs used up until about 1976, Dave.


Tangerines;

Metal phase plugs, remember, never take a driver apart like I did, it must be recharged before re-use;

 

[Zilch]

Zilch, my existing crossover is factory Figure 3 attached .pdf. Drivers are 8 ohm. Please link me to correct crossover mods for group buy 4552ND drivers.

Thank you for posting the AS-101 schematic. The supplemental compensation circuit is here:

http://audiokarma.org/forums/showthread.php?p=1141355&#post1141355

Dgwojo has answered your other questions. Thanks, Dave, for the cool driver dissection pics!

This thread is about 811B horns, generically, and what drivers might be used with them, so we're discussing a variety of options of different cost and quality. I'm just testing some of them that members are sending me on 811B, for the edification of any and all who might be interested.

None of this constitues an analysis of their suitability for any particular application. However, the BMS drivers have been successfuly used by several members and others for upgrading vintage Altec systems.... :thmbsp:

 

[rustnrot]

Earlier in this thread was mention of the Selenium D220Ti. Here are response curves from the. pdfs for this and the D2500Ti-Nd (the only way I knew how to capture an image from the .pdf and convert to .jpg was print-and-scan)

It appears that the 220Ti has even flatter response characteristics than the BMS 4552ND driver. Neither of these drivers exhibit typical horn rolloff it looks like. Also for comparison I have posted the 4552ND.

I am thinking of using the 220Ti driver (crossed at 1500 Hz--the info on PE that says minimum cross at 2000 Hz is incorrect as 1500 Hz is what is in the .pdf).

Since this driver does not seem to need any compensation, what would be "wrong" with a brand new 12db/octave 1500 Hz two-way crossover in my Heathkit AS101/Valencias with a simple L-Pad for the HF horn attenuation?

Comments please.

 

[Zilch]

Since this driver does not seem to need any compensation, what would be "wrong" with a brand new 12db/octave 1500 Hz two-way crossover in my Heathkit AS101/Valencias with a simple L-Pad for the HF horn attenuation?

Note that the Selenium response curves are measured on a Selenium model HL14-25 horn. As I mentioned earlier in this thread, that is a narrow-dispersion exponential horn which compensates for the compression driver rolloff above mass breakpoint acoustically, and is used to show the response in its best possible light.

The only way to know how those drivers will perform on 811B horns (or any others, for that matter,) is to measure, which I will do later this week, once the D220Tis arrive here. While the curves do look better than those for D210Ti, I can virtually guarantee that some HF compensation will still be required.

The BMS curves, on the other hand, are shown measured on a constant directivity horn, which exhibits the characteristic rolloff, and reciprocally, the requisite compensation, in full measure. I have previously posted how BMS drivers actually measure on 811Bs, ad infinitum and ad nauseum.... :

http://audioheritage.org/vbulletin/showthread.php?t=4410

Note: Your proposed move to 1.5 kHz assumes the woofer plays well up to there, and that the stock crossover is not supressing any response anomalies above the crossover frequency. That's nearly an octave higher than AS-101's 800 Hz. It may be fine, but that should be verified via measurements.

 

[rustnrot]

Excellent, I will be standing by waiting for results of your tests, any satellites that may be falling from the sky also later this week, etc....

 

[macaltec]

I just got around to trying out the D210Ti's this evening. On a rudamentary XO setup at best they sound pretty good to me. Certainly not any worse than my current Altec and Eminence offerings. I'm hoping the Zilchster has got some hop up recommendations for dialing them in a bit more (passively).

 

[Zilch]

I'm hoping the Zilchster has got some hop up recommendations for dialing them in a bit more (passively).

See posts #53 and #78.

30923 T-filter with R2 = 15 Ohms works (R1 = 4 Ohms in lieu of 3.9.) Just add it on after your highpass filter.

Simple 6uF with 8 Ohms, effectively 30923 without any R2, works too.

Model 19 crossover works a bit better; the contour is different.

I didn't drag out the JBL 3120A; I'll do that for the D220Tis, which are on the way, as another option, as would be the Pi Speakers compensation circuit, which I haven't tried.... :yes:

 

[skywave-rider]

Great thread! 4552NDs, eh?

 

[Zilch]

HI, Skywave! :wave:

For those not following the unfolding 9844-8B saga over on LHF, Skywave's presently building some modified Model 19 crossovers for use with 902-8B drivers on 811B horns over there. Those are mated with a pair of 12" 414-16E woofers in his pro monitors, similar to what's in the familiar consumer Carmel and Malibu systems.

 

[geaugafletcher]

Note: Your proposed move to 1.5 kHz assumes the woofer plays well up to there, and that the stock crossover is not supressing any response anomalies above the crossover frequency. That's nearly an octave higher than AS-101's 800 Hz. It may be fine, but that should be verified via measurements.

It's been said in hushed, reverent tones that the 416 will work just fine up to 1200Hz - that in fact, the 806/811 and 416 combo sounds better crossed-over at 1200 than 800.

Rumor only, you understand.

I ran into an Altec 1632A active the other day. It has an experimenter's delight for a setup when it comes to horn HF compensation. A little research turned up the manual with all the details. Three resistors to set the frequency and 'height' of the boost, two more resistors as optional (and adjustable) notch filter. Should be fun!

 

[Zilch]

Rumor only, you understand.

I believe it's generally acknowledged, at least based upon anecdote, that 811B is a better 1200 Hz horn than an 800 Hz one. The best evidence is Model 19.

Examined in detail, it also becomes apparent that in many of the earlier products such as Valencia the highpass crossover is NOT electrically 800 Hz, but rather, considerably higher, and provides a measure of compensation for the 806's aggressive midrange response. It's the acoustic response which defines the actual crossover frequency, not the electrical drive alone.

What I'm saying is, short of measuring the acoustic response in the system, we don't know where that exactly is, and must rely upon whatever Altec determined it to be in the specs. I have no data on 416 variants other than electrical sims of Altec crossovers and box models based upon T/S parameters.

I ran into an Altec 1632A active the other day. It has an experimenter's delight for a setup when it comes to horn HF compensation. A little research turned up the manual with all the details. Three resistors to set the frequency and 'height' of the boost, two more resistors as optional (and adjustable) notch filter. Should be fun!

Yes, I've certainly observed the effect of different compensation contours on frequency response, and the optimum contour is specific for each driver/horn combination. JBL, for example, in its early active crossovers, provided different custom plug-in contour shaping cards for it's different families of Biradial horns, flat front vs. constant coverage, as well as for specific systems. In later versions, contours were switch-selectable, and most recently, parametrically defined as "Tunings...."

 

[GordonW]

Examined in detail, it also becomes apparent that in many of the earlier products such as Valencia the highpass crossover is NOT electrically 800 Hz, but rather, considerably higher, and provides a measure of compensation for the 806's aggressive midrange response. It's the acoustic response which defines the actual crossover frequency, not the electrical drive alone.

IME, this is correct. In working with the Heathkit AS-101s here that belong to my friend, I noted it DID have a lower-Q, higher-frequency highpass (smaller cap) than it would need, to be a "textbook" matching-values crossover (between lowpass and highpass).

And, in fact, I wound up using an EVEN SMALLER cap (further pushing up the frequency and lowering the Q), which made the system even more "listenable". Anyone who's got AS-101s, you might want to try a 4.7uf of 6.8uf cap on the horn, instead of the factory 8uf value... IME, it really helped.

That, and adding the HF compensation circuit Zilch posted, can make a nearly night-and-day difference, IMHO. We got this one to that "toe-tapping" point, where we just about couldn't drag ourselves away from listening to them, afterwards!

On another note- I'm about to start building another system, in similar scope to the ones mentioned. I was given a set of "Chrysler" branded cabinets, that were OBVIOUSLY made by Coral... 5.25 cubic feet, made for a 15" woofer. I'm going to re-cut a new baffle board, to use a 15" driver (haven't decided on whether it will be a reconed EV SRO 15 or a reconed Utah 15- have to listen to both in the cabinet), and I'm definitely using the Selenium D210Ti on the same horn as I used in the Seeburgs. "Baby Seeburg", indeed!

Got the critical parts already (cabinets, horns and woofer frames)... now it's just a matter of ordering the rest of the stuff (horn drivers, crossover components, woofer recone kits)...

Regards,
Gordon.

 

[skywave-rider]

4552nd

I'm joining the group buy, if I'm not too late. My motivation is to see if ironing out the roller coaster high end I get from 902s on 811Bs makes a difference. At this time I have no plans to sell the 902s. Zilch, if you need to measure some, let me know.